Previous studies of media effects have been primarily concerned with whether television affects children's social behavior and have neglected to investigate 1. processes by which effects occur and 2. developmental changes in the nature of effects. A series of laboratory and field experiments is proposed relevant to the general hypothesis that effects of complex television presentations are mediated by a cognitive state in the observer. The content of this mediator depends in part upon abilities in the observer that change developmentally; e.g., selective attention, judgments about social acts, and inferences about causal relationships. The specific objective of the proposed studies is examination of age-related differences in comprehension and evaluation of aspects of television content that are potentially relevant to behavioral effects. Children from the age range 6-16 years will view edited television programs and videotapes that vary in the extent of cognitive processing required for comprehension. They will be interviewed according to standardized schedules and will respond to multiple-choice and other structured tasks regarding their recall of and inferences about the programs, particularly behavior-relevant cues like motives and consequences for a depicted act. The instruments and tasks will be developed from extensive pretesting with adult judges and children of different ages, as well as from detailed structural analysis of the narrative itself. In some studies, evaluation and behavioral effects will also be assessed. The studies are relevant to existing theory and data in the areas of personality, socialization, cognitive development, and mass media effects.